MATHEMATICAL CHARACTERIZATION OF INTERFACE INTERACTION NETWORKS

Interface Interaction Network;IIN;random network;random IIN;Applied Mathematics & Statistics

Shapiro, Benjamin HJohnson, Margaret E. ;

Johns Hopkins University

关键词: Interface Interaction Network;    IIN;    random network;    random IIN;    Applied Mathematics & Statistics;   

Others  :  https://jscholarship.library.jhu.edu/bitstream/handle/1774.2/59361/SHAPIRO-THESIS-2016.pdf?sequence=1&isAllowed=y

瑞士|英语

来源: JOHNS HOPKINS DSpace Repository

【 摘 要 】

Biological systems carry out complex functions such as DNA transcription, cell division, and signaling within and between cells through the collective effort of a diverse range of proteins. In these functional pathways, proteins both cooperate and compete with one another to bind their partners, such that the outcome depends on the protein concentrations, regulation of binding sites, and the number of partners per protein. A tool to help scientists visualize some of these interdependences is known as a protein-protein interaction (PPI) network. In these networks, nodes represent proteins and a line is drawn between two nodes if the proteins interact. These networks have allowed for the visualization and analysis of complex systems yet still fail to accurately capture the competition between proteins. Interface-interaction networks (IINs) are a modification of PPI networks that capture such competition. These networks are distinctive because they are constrained by the parent PPI network, and they therefore have novel attributes that have not been previously characterized theoretically. In these networks, nodes represent binding sites and two nodes are connected if the binding sites interact. The structure and topology of the IIN may reflect evolutionary pressures on individual proteins because proteins evolve to recognize and bind their functional partners in specific ways. If the IIN is under evolutionary pressure, it should exhibit unique structural properties. To recognize unusual and unexpected features of a network, we must understand the topology of these networks that would arise under purely random conditions, which will be the goal of this thesis. We will begin by rigorously defining the space of all possible IINs and then enumerate such a space.We will find the number of possible networks to be on the order of 1E177.Using this knowledge, we will develop a statistical test to determine if an IIN;;s topology is under selective pressure. This test will be applied to the clathrin-mediated endocytosis (CME) IIN in yeast and will allow us to conclude that this network is under selective pressure.Finally, we will characterize the global structural properties of the random networks and compare them to the CME IIN. We will find that the CME IIN has a unique scale free distribution. We will conclude with a brief discussion on local motifs and the difficulties involved with delineating their prevalence in IINs.